Imagen 4 Ultra
Imagen 4 Ultra is Google's maximum-fidelity image generation model, the quality ceiling of the Imagen 4 family, designed for print production, publication-grade visuals, high-DPI displays, and professional creative applications where every detail in the rendered image matters.
import { experimental_generateImage as generateImage } from 'ai';
const result = await generateImage({ model: 'google/imagen-4.0-ultra-generate-001', prompt: 'A red balloon on a wooden table.'});Playground
Try out Imagen 4 Ultra by Google. Usage is billed to your team at API rates. Free users (those who haven't made a payment) get $5 of credits every 30 days.
About Imagen 4 Ultra
Imagen 4 Ultra is the highest-fidelity tier in the Imagen 4 family. It launched alongside the fast and standard variants as part of AI Gateway's image-only model category, purpose-built for professional and creative applications where standard-tier image quality doesn't meet the bar.
The distinction between ultra and standard is not primarily about resolution. It's about what happens in those pixels. Ultra-tier generation emphasizes fine detail rendering: fabric texture, surface reflections, fine hair and skin detail, material microstructure, and depth-of-field rendering in complex scenes. In a web thumbnail, these differences may be invisible. In a full-bleed print spread, a billboard, a publication hero image, or a product shot viewed at full zoom, they determine whether an image looks generated or holds up to professional scrutiny.
For design agencies, advertising studios, and media production teams, the ultra tier serves as the final-output generator in a pipeline that begins with fast-tier iteration. You explore prompts quickly, select candidates, and submit only approved prompts to ultra for final rendering. This keeps the bulk of generation cost on the fast tier while ensuring final output reflects Imagen 4's maximum capability.
Print production is a particularly clear use case. Images destined for physical print media require detail density that digital-only assets do not. Ultra-tier generation provides the photorealistic quality floor that makes AI-generated images viable in professional print contexts.
Providers
Route requests across multiple providers. Copy a provider slug to set your preference. Visit the docs for more info. Using a provider means you agree to their terms, listed under Legal.
P50 throughput on live AI Gateway traffic, in tokens per second (TPS). Visit the docs for more info.
P50 time to first token (TTFT) on live AI Gateway traffic, in milliseconds. View the docs for more info.
More models by Google
| Model |
|---|
| 1M | 0.9s | 317tps | $0.25/M | $1.50/M | Read:$0.03/M Write:— | $14.00/K + input costs | — |   | 03/03/2026 | ||||
| 1M | 3.2s | 147tps | $2.00/M | $12.00/M | Read: $0.2/M Write: — | $14.00/K + input costs | — | | 02/19/2026 | ||||
| 1M | 0.6s | 187tps | $0.50/M | $3.00/M | Read: $0.05/M Write: — | $14.00/K + input costs | — | | 12/17/2025 | ||||
| 1M | 0.3s | 283tps | $0.10/M | $0.40/M | Read:$0.01/M Write:— | $35.00/K + input costs | — | | 06/17/2025 | ||||
| 1M | 0.3s | 207tps | $0.30/M | $2.50/M | Read:$0.03/M Write:— | $35.00/K + input costs | — | | 03/20/2025 | ||||
| 1M | 1.8s | 127tps | $1.25/M | $10.00/M | Read: $0.13/M Write: — | $35.00/K + input costs | — | | 03/20/2025 |
What To Consider When Choosing a Provider
- Configuration: The ultra tier carries a higher per-image cost than the fast and standard tiers. The practical workflow for most teams is to develop and validate prompts using
google/imagen-4.0-fast-generate-001, then route finalized prompts to ultra for production output; only the model slug changes between phases. - Zero Data Retention: AI Gateway supports Zero Data Retention for this model via direct gateway requests (BYOK is not included). To configure this, check the documentation.
- Authentication: AI Gateway authenticates requests using an API key or OIDC token. You do not need to manage provider credentials directly.
When to Use Imagen 4 Ultra
Best For
- Print production and publication: Images destined for physical media, large-format output, or editorial use where maximum detail density is required
- Hero and brand photography: Images that will be viewed at full resolution where detail quality reflects on the brand
- High-DPI display assets: Premium applications, product pages, and marketing materials viewed at retina resolution
- E-commerce product photography: Fine material detail directly influences purchase decisions
- Final render stage: Pipelines where prompt development already happened on the fast or standard tier
Consider Alternatives When
- Prompt development phase: You are still iterating on prompts (consider
google/imagen-4.0-fast-generate-001to reduce cost) - Web-scale display output: Standard-tier quality is sufficient for standard display resolutions (consider
google/imagen-4.0-generate-001) - Text output alongside images: Mixed-modality generation is required (consider
google/gemini-3-pro-image) - Video output required: Still images are not sufficient (consider the Veo model family)
Conclusion
Imagen 4 Ultra is the right choice when image output quality is load-bearing: when the image will be printed, published, displayed at high resolution, or evaluated by a professional who knows what AI-generated images look like at scale. For those applications, the ultra tier's quality ceiling is what justifies its place in the Imagen 4 lineup.
Frequently Asked Questions
What does "ultra" fidelity mean concretely, and how is it different from standard?
The ultra tier produces richer detail in fine textures, surfaces, and complex lighting scenarios. Material microstructure, depth-of-field rendering, and fine subject detail are rendered with greater precision. At web thumbnail sizes, this may be invisible. At print resolution, on high-DPI displays, or in zoomed product imagery, the ultra tier holds up to scrutiny that standard-tier images may not.
Is the ultra tier designed for print production?
Yes. Print media requires higher detail density than digital-only assets. Ultra-tier generation provides the photorealistic quality floor that makes AI-generated images viable for full-bleed spreads, large format, and editorial use.
How do I justify the ultra tier's cost for my use case?
The cost is justified when image quality has direct business value and lower fidelity would make an image unsuitable for its intended use. For a thumbnail in an internal tool, standard tier is sufficient. For a hero image in a brand campaign or a product shot in a premium catalog, the ultra tier's quality is part of the deliverable.
What is the recommended workflow for cost-efficient use of the ultra tier?
Develop and validate prompts using
google/imagen-4.0-fast-generate-001. When prompts produce the correct subject, composition, and style, run the finalized prompts throughgoogle/imagen-4.0-ultra-generate-001for final output. The AI Gateway unified API means only the model slug changes between iteration and final-render phases.Does the ultra tier produce images at a higher resolution than standard?
Ultra generation targets maximum rendered detail and fidelity within the supported resolution parameters. For exact resolution specifications, see the Specs table above. The quality difference is in rendering detail rather than pixel dimensions.
Does this model accept image inputs for editing?
Imagen 4.0 Ultra is a text-to-image generation model. For image editing and inpainting use cases, check the AI Gateway model catalog for models that support those workflows.
How do I call Imagen 4.0 Ultra from the AI SDK?
Use
experimental_generateImage(aliased asgenerateImage) withmodel: 'google/imagen-4.0-ultra-generate-001'.Do I need a Google Cloud project to access this model?
No. AI Gateway manages all provider credentials. Authenticate using a Vercel API key or OIDC token.